A Comparative Study of Lexical Stress Placement in the Speech of Taiwanese ESL Learners and American English

Abstract

It is well known that English is a stress-timed language. Furthermore, it is believed that the English spoken by Taiwanese ESL learners is not. This paper compares the stress characteristics of Taiwanese speech with that of American English through acoustical examinations. The study contrasts the realization of stress placement in two intonation-phrasal conditions in the two speech groups. Acoustical results indicate that the lengthening of the final syllable of multi-syllabic words at the phrase-final condition does not occur as it does in Singaporean English. The lexically unstressed syllables following the primary stress in both speech groups show a steady decrease in duration. The overall f0 movement is similar for both speech groups and shows no shift of prominence, with the Taiwanese group displaying a richer tonal movement. Contrastively, after the penultimate syllable in the phrase-medial position, the pitch contour is more level for both speech groups. Both sample groups show distinct vowel qualities for all three syllables of the test items. For all stimuli, the Taiwanese group exhibits a higher f1 and a lower f2, which contrasts the American native group that exhibits a higher f2 and a lower f1. Viewed from the f1/f2 formant space, the first syllable is the most contrastive: the Taiwanese first syllables are positioned at the bottom right corner, while the American first syllables are located mid-right in the formant space. The final syllable shows the most distinct patterning, especially within the phrase-medial condition, where f1 and f2 for both speech groups are positioned differently. Syllables spoken by the Taiwanese group are located at the bottom-half of the formant space while the syllables of the American group fall left centre. Overall, Taiwanese ESL learners show distinct differences in realising stress placement of multi-syllabic morphemes with various degrees of stress within the two prosodic structures. Implications for prosodic teaching are also suggested.